Why Your EHR and Billing Reports Never Match

A practice manager pulls the monthly production report from the EHR and compares it against the report the billing company sent over. The two totals do not match, and neither one is obviously wrong. Nobody misentered data as far as anyone can tell, and no system crashed. Report mismatches between an EHR and a billing system are one of the most common data-integrity complaints in outpatient practice, and it is worth understanding why they happen before spending time chasing what looks like a bug.

AAFP's Family Practice Management has documented this exact conflict: a practice's internal EHR reporting and its billing company's reporting can diverge even when both are built from the same underlying claims. The two systems are not designed to answer the same question, and a mismatch between them says more about how each system defines its numbers than about which one made a mistake.

A structural problem, not a data error

An EHR is built around clinical documentation and the encounter record. Its reporting module usually totals charges as they are entered, the amount billed at the code level at the moment a note is closed and coded. A billing system, whether it runs inside the practice or at an outside revenue-cycle vendor, is built around the claim lifecycle: submission, adjudication, payment posting, adjustment, and secondary billing. Its reports describe where claims sit in that lifecycle, not the encounters that generated them.

Run a report from each system for the same month and the two are answering different questions by design. An EHR report usually answers "what did we bill." A billing system report might answer "what did we collect," "what is still outstanding," or "what did the payer allow," depending on which report gets pulled. Compared side by side under one label such as "revenue," with no note on which definition each number uses, the two totals look like an error. They are two answers to two different questions.

Four places the gap usually comes from

No shared definition of the number

The word "revenue" gets used loosely. It can mean charges billed, the allowed amount under a payer contract, or payments actually posted, three figures that can differ from each other by a wide margin. A practice quoting monthly revenue from the EHR is often quoting gross charges. A billing company quoting the same term is often quoting net collections after adjustments and write-offs. Both figures are correct for what they measure. Without a label attached, comparing them looks like a discrepancy that needs an explanation.

One-way interfaces

Most practices run an interface that pushes charges from the EHR into the billing system in one direction only. A biller who corrects a coding error, adds a modifier, or voids a duplicate charge has changed the number inside the billing system. The EHR record that generated the original charge usually keeps the uncorrected version, because the interface was never built to send corrections back. Months of small corrections made only on the billing side, none of them mirrored into the EHR, build into a gap between the two systems that has nothing to do with either being wrong at the point of entry.

Timing lag between systems

The two systems rarely close a reporting period at the same moment. A claim entered in the EHR on the last day of the month may not appear in the billing system's export until an overnight batch runs, which can push it into a different reporting month entirely. Add the normal delay between claim submission and payer adjudication, often several weeks, and a report pulled from each system on the same calendar day can accurately reflect two different sets of claims.

Manual re-entry and drift

Where systems do not interface at all, or the interface covers only part of the workflow, someone is retyping charges, payments, or adjustments from one system into the other by hand. Each manual entry point is a chance for a transposed digit, a skipped line, or a charge logged under the wrong date of service. No single error looks significant. Accumulated over months across hundreds of line items, they settle into totals that no longer agree, and by the time anyone checks, there is no single mistake to point at, only accumulated drift.

Finding out which cause is yours

Before treating either report as wrong, work through the mismatch in order:

• Write down, in one sentence, what each report actually counts: charges entered, the allowed amount, or payments posted. A different definition in each system can explain most of the gap by itself.
• Check how each system defines the reporting period: date of service, date of entry, or date of payment. A month boundary that means something different in each system will produce a mismatch on its own.
• Pick one claim that appears differently in each system and trace it from entry to posting: date entered, date submitted, date adjudicated, date paid. A gap that closes once the claim finishes its lifecycle is a timing lag, not an error.
• Confirm whether corrections made in the billing system get sent back to the EHR, or only flow in one direction. A one-way interface widens the gap a little more every month.
• Map every point where a person retypes a number from one system into the other, and check whether that step has been audited recently for accuracy.
• Once the cause is identified, name one system as the system of record for that specific number, in writing, and treat future comparisons as reconciliation, not as a search for which report is broken.

When the gap deserves more attention

A small, stable difference between two reports that measure different things is expected and does not need fixing. A gap that keeps growing month over month, rather than settling out once claims finish adjudicating, usually points to a real process problem: an interface that never sends corrections back, or a manual step nobody has checked in years. Working through the causes in order replaces a guess about which report is right with a specific, checkable answer, and that answer is almost always fixable once someone has named it.